Analysis

Neuromorphic computing a potential reality with atomic switch

1st August 2014
Staff Reporter
0

The mammalian brain continues to reign supreme in completing tasks based on both memory and environmental clues, even though computers have revolutionized information processing. This contrast in performance stems from the radically divergent physical structures and operating mechanisms of neuronal networks and digital circuits.

 

Computers employ a microprocessor to rapidly perform simple, error-free calculations in a sequential fashion and store data in physically separate memory banks. In contrast, the brain comprises a vast network of neurons serving simultaneously as both information processors and memory units, resulting in comparatively slow and imprecise operations in a parallel or distributed manner.

In the past, efforts to mimic brain function involved programming computers to create virtual neural networks. However, a neuromorphic device designed to incorporate structural aspects inspired by the cortical neuropil is being developed at the California NanoSystems Institute (CNSI) and the International Center for Materials Nanoarchitectonics (MANA). The aim of the neuromorphic device is to produce the class of operational properties which underlie cognition in the mammalian brain.

The atomic switch, a recently developed nanoscale circuit element, has been shown to possess synapse-like properties in a purely inorganic device. Using a nanoarchitectonic approach, millions of atomic switch elements are incorporated into a densely interconnected network of silver nanowires. These atomic switch networks (ASN) retain the synaptic properties of their of individual component elements and generate emergent behaviors comprised of their distributed, collective interactions. Such emergent behaviors are a principal characteristic of biological neural networks and many other complex systems. Ongoing studies involve the utilisation of these emergent behaviors for information processing toward the generation of a new class of cognitive technologies.

Inside the ASN device is a highly interconnected architecture which comprises synaptic circuit elements at each point of contact between nanowires. The collective interactions between these atomic switches result in unique, emergent properties which have shown significant potential for neuromorphic computing.

Product Spotlight

Upcoming Events

View all events
Newsletter
Latest global electronics news
© Copyright 2024 Electronic Specifier